Hi, I'm Manideepu Reddy Enugala 👋

I'm a data-driven machine learning developer with a strong focus on Generative AI, Machine Learning, Deep Learning and Computer Vision. Every day, I thrive on solving complex problems, refining AI models, and pushing the boundaries of financial technology with innovative solutions."

Hyderabad, Telangana

Available for new projects

About me

Curious about me? Here you have it:

I'm a passionate, self-proclaimed AI developer who specializes in Generative AI and Computer Vision. I am enthusiastic about reading up on new technologies in AI and understanding the mathematics behind algorithms. I have a strong foundation in machine learning, deep learning, and computer vision, and I love to apply these skills to solve real-world problems.

My interest began in college in 2022, and since then, I've continued to learn and evolve, taking on new challenges and learning the latest technologies along the way. I always wanted to become a Robotics AI developer.

In my free time, I enjoy working on personal projects, contributing to open-source, and exploring new ideas in the field of AI.

Finally, some quick bits about me:

B.Tech. in Mechanical Engineering
Part time freelancer
Avid learner
I love to build things

Skills

The skills, tools and technologies I am really good at:

Git

Python

SQL

PyTorch

Machine Learning

Hugging Face

Azure

Docker

Langchain

Experience

Here is a quick summary of my most recent experiences:

Data Scientist

Innovated a deep learning framework for fraud detection, enhancing anomaly identification and improving risk mitigation processes.
Designed a machine learning model for SME Working Capital loan disbursals, automating decision-making and reducing bad rate of loan approvals.

Aug 2024 - Present

Work

Some of the noteworthy projects I have built:

Research Project, Prof. Sankha Deb, IIT Kharagpur

Integrated RealSense camera and detectron for precise point cloud segmentation in cluttered spaces. Designed 3D registration techniques to determine object pose and position for robotic manipulation. Engineered a two-finger linear adaptive gripper for optimal grasping with advanced trajectory planning.

RealSense

Detectron

Point Cloud

3D Registration

Robotics

Trajectory Planning

Fine-Tuning LLaMA-2 for Medical QnA with QLoRA

Utilized QLoRA to fine-tune Llama-2-7b on Medical QnA, optimizing for medical QA. Used 4-bit precision, LoRA parameters, gradient checkpointing, and learning rate scheduling for efficient training on limited resources.

LLaMA-2

QLoRA

HuggingFace

4-bit Training

Gradient Checkpointing

Medical QnA

Fine Tuning Diffusion Model on Emoji Dataset

Implemented and fine-tuned a DDPM-based generative diffusion model on an emoji dataset, using advanced preprocessing and training strategies. Improved inference speed with DDIM and deployed on HuggingFace Hub.

Diffusion Model

DDPM

DDIM

HuggingFace

Image Preprocessing

Multi-Source RAG Chatbot

Deployed a Streamlit web app integrating OpenAI LLMs and RAG. Used data extraction for PDF, TXT, CSV, stored in FAISS DB, and Langchain retriever for context-based responses.

Streamlit

OpenAI

RAG

FAISS

Langchain

PDF/TXT/CSV

DCGAN and Conditional AC-DCGAN Model Development

Built a Cat image generator with DCGAN (8M params) on 64x64 cat dataset. Implemented Conditional AC-DCGAN on CIFAR-10 (5 classes) using PyTorch and GPU for fast training.

DCGAN

AC-DCGAN

PyTorch

CIFAR-10

GPU Training

Get in touch

What’s next? Feel free to reach out to me if you are looking for a developer, have a query, or simply want to connect.

enugalamanideepreddy99@gmail.com

9121032689

You may also find me on these platforms!